-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Are we heading for a time when all CDs will be MQA coded - no redbook available?
- Thread starter Burning Sounds
- Start date
I'm guessing they will use Ogg container with Flac encoding. They might have already anticipated this and made sure their device SDK's can already handle this. In that case, my concerns are unfounded.
Last edited:
- Joined
- Oct 25, 2019
- Messages
- 11,098
- Likes
- 14,755
https://www.lifewire.com/what-makes-an-audio-format-lossy-2438561
The interesting bit for me is what induces the strong physical reaction you have to MP3- is it the method they use to do compression or are you averse to lossy files generally, regardless of file type? Rhetorical questions by the way, I know you likely wont have done a detailed exploration.
All internet radio, for example, will be lossy music in one format or another, as will music played on TV, on DVD etc etc. May not all be MP3 specifically though.
Actually, I would argue that MQA needs MORE bandwidth than a comparable non-MQA encoded FLAC variant. For the sake of argument let's just assume that having any audio above 22 kHz is totally useless. This means we would at most need 48 Khz sampling. Now I FLAC encode the 24 bit, 48 kHz master file. I'll also do the same with the MQA encoded version of the same file. I bet you the MQA file is larger than the original master. The reason for that is that it contains a lot more noise. Noise is inherently hard to compress.
You can simply test this. I generated a 30 second 440 Hz wave (0.8 amplitude) in Audacity and saved this as FLAC. The file is 1.3 MB. Now I add 0.001 amplitude of noise to it and save it as a FLAC file again. Now the file is suddenly 2.5 MB in size. And even with 0.0001 amplitude noise, it's still 2 MB! Obviously, this is an extreme example, but with MQA the results it will be equivalent. Surely, the differences will be much smaller though.
You can simply test this. I generated a 30 second 440 Hz wave (0.8 amplitude) in Audacity and saved this as FLAC. The file is 1.3 MB. Now I add 0.001 amplitude of noise to it and save it as a FLAC file again. Now the file is suddenly 2.5 MB in size. And even with 0.0001 amplitude noise, it's still 2 MB! Obviously, this is an extreme example, but with MQA the results it will be equivalent. Surely, the differences will be much smaller though.
Last edited:
Am guessing something has changed in the blueos lib as your work is no longer reproducible, and the rendered file is not playable (though it is 24bit). Starting with the same file.
Code:
$ ./mqascan -1 2L-110_04_stereo.mqa.flac
00000000: MQA signature at bit 8
00000000: [5] datasync
magic 36: 0x11319207d
stream_pos_flag 1: 0
pad 1: 1
orig_rate 5: 0x03 [352.8 kHz]
src_rate 5: 0x00 [44.1 kHz]
render_filter 5: 6
unknown_1 2: 0
render_bitdepth 2: 1 [18 bits]
unknown_2 4: 0x0
auth_info 4: 0x0
auth_level 4: 0x9
item_count 7: 2
size 8: 0x14
size 8: 0x0b
type 8: 0x00
type 8: 0x01
[type 0]
stage2_dither 2: 2
gain_index 4: 0
unknown_5 7: 15
unknown_6 7: 127
[type 1]
unknown_7 6: 25
unknown_8 2: 1
unknown_9 1: 0
unknown_10 2: 1
checksum 4: 0xc
Code:
$ ./mqadec 2L-110_04_stereo.mqa.flac decoded.wav
mqaAuthored/352800
Code:
$ sox decoded.wav -n stats
Overall Left Right
DC offset -0.000000 0.000000 -0.000000
Min level -0.887951 -0.849381 -0.887951
Max level 0.930965 0.909392 0.930965
Pk lev dB -0.62 -0.82 -0.62
RMS lev dB -21.89 -21.80 -21.98
RMS Pk dB -9.90 -10.38 -9.90
RMS Tr dB -69.47 -69.47 -69.06
Crest factor - 11.19 11.69
Flat factor 0.00 0.00 0.00
Pk count 2 2 2
Bit-depth 24/24 24/24 24/24
Num samples 30.2M
Length s 342.880
Scale max 1.000000
Window s 0.050
Code:
$ ./mqarender -r 352800 decoded.wav rendered.wav
Code:
$ soxi decoded.wav rendered.wav
Input File : 'decoded.wav'
Channels : 2
Sample Rate : 88200
Precision : 24-bit
Duration : 00:05:42.88 = 30242016 samples ~ 25716 CDDA sectors
File Size : 181M
Bit Rate : 4.23M
Sample Encoding: 24-bit Signed Integer PCM
Input File : 'rendered.wav'
Channels : 2
Sample Rate : 352800
Precision : 24-bit
Duration : 00:05:42.88 = 120968064 samples ~ 25716 CDDA sectors
File Size : 726M
Bit Rate : 16.9M
Sample Encoding: 24-bit Signed Integer PCM
Total Duration of 2 files: 00:11:25.76
Code:
$ sox rendered.wav -n stats
Overall Left Right
DC offset -0.000001 -0.000001 -0.000001
Min level -0.887941 -0.849387 -0.887941
Max level 0.930983 0.909385 0.930983
Pk lev dB -0.62 -0.83 -0.62
RMS lev dB -27.91 -27.82 -27.99
RMS Pk dB -14.76 -15.34 -14.76
RMS Tr dB -76.78 -76.78 -75.40
Crest factor - 22.38 23.37
Flat factor 0.00 0.00 0.00
Pk count 2 2 2
Bit-depth 24/24 24/24 24/24
Num samples 121M
Length s 342.880
Scale max 1.000000
Window s 0.050Though the rendered file is not playable, its all glitches/noise. The decoded file is still playable.
Edit. Actually it does render if you don't specify the rate. This produced the full file. Perhaps there was simply a bug in earlier versions of the lib which lead you to believe it wasn't doing what MQA claimed?
Code:
$ ./mqarender decoded.wav rend.wav
Code:
$ sox rend.wav -n stats
Overall Left Right
DC offset -0.000000 0.000000 -0.000000
Min level -0.890860 -0.849512 -0.890860
Max level 0.930959 0.909541 0.930959
Pk lev dB -0.62 -0.82 -0.62
RMS lev dB -21.89 -21.80 -21.98
RMS Pk dB -9.90 -10.38 -9.90
RMS Tr dB -69.45 -69.45 -69.04
Crest factor - 11.19 11.69
Flat factor 0.00 0.00 0.00
Pk count 2 2 2
Bit-depth 24/24 24/24 24/24
Num samples 60.5M
Length s 342.880
Scale max 1.000000
Window s 0.050
Last edited:
The "decoded" file is bit for bit identical with new and old versions of the library. The rendering instructions are in the LSB, so it will always appear to be 24-bit in a trivial analysis.
You're right, there seems to be an error somewhere.
Looking closer and it seems without specifying the rate, mqarender actually "rendered" a 176.4 kHz file not 352.8. ~60M samples instead of ~120M, though it is playable. Didn't have a chance to look at the code, technically supposed to be working right now ;-)
Yes, it works for 2x the input rate. As I'm also "working," I'll look into it later.
I listen to over the air FM as far as radio is concerned (2012 Lexus ES350 with Mark Levinson sound system [yes, I know that's Harmon and Mr. Levinson's history] or (at home) an NAD 4300 AM/FM tuner. All I know is that in 2002 some "mp3's were given to me to listen to with headphones, I had a physical reaction (severe headache) after 3 songs. I tried a couple more times shortly after that with MP3's, decided that MP3's are not for me. And never listened to any that I know of again. CD's, DVD's Blu Ray, 4K, 8K,. FLAC, LP's & 78's, R2R (oddly, one of my 2 AKAI R2R's has an 8 track built into it), cassette. Of these things, only the mp3's bothered me (through headphones). Perhaps it was low quality, not knowing anything about it (and still don't}, I damned MP3 & never knowingly listened to any MP3's again
Don Hills
Addicted to Fun and Learning
Almost all radio stations play music from compressed files on servers...
MP3 bitrates and encoder quality have come a long way since 2002.
MP3 bitrates and encoder quality have come a long way since 2002.
restorer-john
Grand Contributor
Our local radio stations were apparently going to re do all their lossy files with uncompressed, but the big issue was the original CD content was long gone. Clearly, based on the few times I have a quick listen, it's just as bad, if not worse. I can't listen to FM radio anymore, tracks/songs/music I know are just butchered by the coding/compression and artefacts.
If I listened (and or watched) during those years, it was through my ADVENT 300 receiver, Kenwood cassette deck, Connoisseur TT, & any of a # of vcr's, & a couple of different CD players and Frazier Super Monte Carlo speakers. In 91 I switched to a DENON 5.1 setup which, when I had everything set up, I did not think much of it. I also obtained Dahlquist M-905 speakers. I got rid of the DENON & kept both the Frazier's and the DQ's (both of which are in current use. But it was rare (& still is) for me to sit & listen to music or watch a TV program (maybe a movie some times [3 or 4 a year]). I generally have many things that take precedence over that. The older I get, the more things there are to get done.Nah, you're not getting off that easily... The real question is: were have you been from 1993 and 2001
Fortunately a couple of the stations here seem to sound good. Most do not sound good. I think that they figure that people in cars don't deserve good quality.
Spotify (any streaming service) must use the music files provided by the record labels. There is a real possibility that Warner Music Group might give Spotify the same 16/44.1 MQA encoded files they give to Tidal.
Zensō
Major Contributor
Let’s hope not, but I agree, it’s a real possibility.
That may be but I have FM. To my knowledge, I have no MP3's of any type. I listen to FM because I like the programing on the particular stations I listen to. It is certainly not for 'critical listening'. But it suffices for background noise.
I guess that everyone that mentions the MP3 format to me thinks that I am listening to FM only for music. There are books being read, discussions happening, news, weather, traffic, gardening, etc and I get all of these thing from a hundred mile radius (with my antenna height). As I said, it is the programming, some of which is music. I don't turn a TV on, as that requires too much of my attention.
It compiles fine once I install the libgmp-dev package. Are there usage instructions anywhere?
